Burning down the house: Thymic repair and regeneration after acute damage

SUMMARY

The thymus is extremely sensitive to insult but also has a remarkable capacity for endogenous repair. However, even though there is continual thymic involution and regeneration in response to everyday insults like stress and infection, profound thymic damage such as ionizing radiation leading to prolonged T cell lymphopenia for which there is currently no therapeutic treatment. We and others have been focusing in recent years on untangling the cellular and molecular mechanisms underlying endogenous thymic regeneration in the hope of being able to exploit them for clinical benefit. To date multiple molecular mechanisms have been identified that are centered on several distinct cell axes, including Interleukin-22 produced by innate lymphoid cells, BMP4 by endothelial cells, and Type 2 cytokines from eosinophils, ILCs and Tregs. Notably, one of the uniting triggers for these pathways of repair center on the balance of cell death detection. IN this review we will highlight the current state of play with regard to cellular and molecular pathways of regeneration as well as the mechanisms triggering them. We will also highlight recent work that sheds light on the limitations of thymus repair and speculate as to what will be needed for an effective thymus-boosting therapy.

1. INTRODUCTION

Despite its importance for generating and maintaining a broad and self-tolerant T cell receptor repertoire, thymic function is highly dynamic, changing both acutely and chronically in response to stressors. Notably, this dynamism in the tissue has been known for far longer than its role in immune function, where in one of the thymus’ first recorded descriptions, Greek physician Galen (129–216AD) noted its decline in size with age (1, 2). However, the capcity of the thymus to regenerate has also been noted for over a century, with reports that adrenalectomy and gonadectomy lead to significant thymic regrowth (3, 4). In the surgical setting, where thymectomy is routinely performed during pediatric cardiac surgery, children who have undergone partial thymectomy exhibit significant rejuvenation of the remaining thymic tissue (5). In addition to this progressive decline in thymic function with age, the thymus is also hyper sensitive to stimuli that lead to acute involution, including everyday insults like stress, infection, trauma, drug use, pregnancy, malnutrition, and chemical exposure, as well as more profound damage such as cytoreductive chemotherapy, ionizing radiation, and malnutrition (6). However, despite this sensitivity, the thymus also has a remarkable capacity to repair itself after injury. Endogenous thymic regeneration is therefore a crucial – and evolutionarily conserved – process that ensure restoration of immune competence after acute insult (6). Much of what we know about thymic regeneration clinically after acute insult comes from studies in allogeneic hematopoietic cell transplantation (allo-HCT), which sees considerable thymic damage caused by the myeloablative cytoreductive conditioning required to achieve successful hematopoietic engraftment (7, 8). Consistent with evidence that constriction of the TCR repertoire with age leads to poor immune responses, prolonged post-transplant lymphopenia, which can be largely attributed to reduced de novo T cell production due to poor thymic function, is associated with increased risk of infection, relapse and formation of secondary malignancies (7, 8). Not surprisingly, recipient age is a major predictor of T cell reconstitution and clinical outcomes (9–12). In fact, there is an emerging body of work to suggest that thymic function and CD4 T cell reconstitution correlates strongly with transplant outcomes, including overall survival, GVHD, and relapse (13–16) (Figure 1). Thymic regeneration is therefore a crucial process that restores immune competence after acute stimuli but could also be harnessed to improve immune function in individuals whose immune system has deteriorated, either due to therapeutics such as chemotherapy and radiation, or as a result of age. However, despite its importance, the cellular and molecular mechanisms that mediate this tissue repair have only begun to be elucidated over the last decade. In this review we will outline the work that has gone into outlining these pathways and their potential for therapeutic translation.

Figure 1: Thymic function and T cell repertoires across lifespan and after damage.

Thymic function at baseline generates a broad and self-tolerant repertoire of T cells but this function gradually declines over the lifespan, leading to both reduced repertoire breadth and decreased tolerance. Acute damage - which can be caused by everyday insults like stress or infection, as well as severe damage such as cancer chemotherapy or ionizing radiation – leads to a profound immediate decline in thymic function leading to the generation of even fewer T cells. The extent and duration of that decline is largely dependent on the damage itself but in general the thymus will recover in function.

2. EXOGENOUS FACTORS PROMOTING THYMIC FUNCTION

Given the tremendous decline in thymic function with age, as well as its importance for recovery of T cell reconstitution after allo-HCT, for many years much of the work in thymus regeneration had been focused on identifying factors that could improve thymic function in either the context acute injury like chemotherapy or allo-HCT conditioning, or in the context of age-related involution (8). However, many of these therapies were identified based on their physiological roles during homeostatic T cell development such as keratinocyte growth factor (KGF), which is involved in TEC ontogeny during development; Interleukin-7 (IL-7), which is a crucial cytokine driving T cell proliferation and differentiation; and providing additional hematopoietic precursor support (8).

Interleukin-7 (IL-7) is likely the most widely studied molecule with thymic regenerative capacity (17), acting on both developing and mature B cells and T cells (17–19). Exogenous administration of IL-7 can enhance thymic function in old mice as well as boosting thymic function after acute injury such as allo-HCT (20, 21). For these reasons there has been considerable interest in developing IL-7 as a therapeutic and studies in patients with either solid tumors or HIV have shown safety as well as efficacy with respect to T cell expansion (22–26). In the context of allo-HCT, we and others have shown that recipients of allo-HCT given a recombinant and glycosylated form of IL-7 increased their peripheral CD4+ and CD8+ T cells and formation of virus-specific T cells (27). However, since IL-7 has significant effects on peripheral T cells directly, many of these effects may primarily be by stimulating peripheral T cells, rather than on the thymus itself (28, 29).

Similarly to IL-7, KGF (or FGF7) has been widely studied for its impacts on improving thymic health. Administration of recombinant KGF significantly increases thymic cellularity in both aged mice as well as those who have undergone acute damage such as allo-HCT conditioning (30–32). However, unlike IL-7, KGF has been approved for the clinical treatment of mucositis (33), making KGF an attractive therapeutic target for improving thymic function. However, while several studies in mice and non-human primates, including our own, have shown efficacy for thymic-dependent T cell development following allo-HCT (34), and several trials have been established to investigate the potential for KGF as a treatment to boost thymic function have been established (NCT01233921, NCT03042585, NCT02356159 and NCT00593554), the one, albeit retrospective trial performed showed no effect of KGF on absolute lymphocyte count after allo-HCT, although no analysis was assessed for CD4 T cell counts or T cell receptor excision circles (TRECs), the gold-standard for measuring thymic function in patients (35). Moreover, a recent trial in patients with multiple sclerosis who had undergone T cell depletion with anti-CD52 demonstrated there was actually a decrease in thymic-dependent T cell development (36).

In contrast to IL-7 and KGF, which have both been well described for their role in T cell development and thymic function, sex steroid inhibition (SSI) is a widely used therapeutic strategy used to treat sex steroid-dependent malignancies such as prostate cancer but is also one of the most potent means of boosting thymic function. In fact, this regenerative capacity of SSI on thymic function has been known since the turn of the last century (37, 38). In a study of prostate cancer patients (who are generally older men who it would be expected to have involuted thymuses) it was found that SSI, using the LHRH agonist Lupron, significantly increased CD4 and CD8 counts as well as in TRECs (39). In a prospective trial of allo-HCT and autologous HCT recipients, SSI enhanced neutrophil engraftment as well as increased TRECs and improve TCR diversity in recipients of both allogeneic and autologous allo-HCT (40). In the thymus, SSI enhances thymic architectural organization as well as the ability to import circulating progenitors (41). This general improved function enhances T cell development and the export of new naïve T cells in mice and humans in both the contexts of aging and damage such as allo-HCT (39, 41–48). While the mechanisms by which SSI can improve thymic function are not well understood, early studies using bone marrow chimeras of sex steroid insensitive mice found that their effects were largely due to the radioresistant thymic stroma rather than the radiosensitive hematopoietic compartment of the thymus (49, 50). Consistent with this, we found that androgens can directly bind and suppress the transcription of the Notch ligand delta-like 4, limiting thymic function, which can be reversed after SSI (51). Inhibition of sex steroids is also known to induce expression of molecules such as CCL25 (41), which is important for the importation of hematopoietic progenitors into the thymus (52, 53). However, in addition to its impacts on the thymic microenvironment, we have also shown that SSI can enhance the function of the most primitive hematopoietic stem cells, likely via modulation of bone marrow niche function (54).

4. UNDERSTANDING ENDOGENOUS MECHANISMS OF TISSUE REPAIR

Innate Lymphoid Cells and Interleukin-22

Despite the extensive work investigating thymus-boosting therapies, little was known how the thymus was capable of robust endogenous regeneration. Given the considerable similarity between the gut and the thymus, both being epithelial tissues that see considerable damage and repair cycles through everyday events as well as more severe insults, we decided to concentrate on a newly identified molecule, interleukin-22, that was being extensively investigated for its role in maintenance of barrier function and induction of innate antimicrobial molecules at mucosal surfaces (55). We had identified that IL-22, produced by innate lymphoid cells (ILCs), acted on gut intestinal stem cells (ISCs) to mediate repair after damage such as that caused by allo-HCT conditioning (56, 57). Furthermore, not only were ISCs a target of GVHD, but so were ILCs and this lack of regenerative cues could impair gut regeneration during GVHD. Notably, although no work had been performed to identify a role for IL-22 in the thymus, the molecule itself was in fact first cloned from thymus tissue (58). Although IL-22 was redundant for steady state thymic development and function, mice deficient for IL-22 exhibited inferior regeneration after sublethal total body irradiation (SLI) compared to WT controls (59).

Much like the gut, we discovered that IL-22 was produced by innate lymphoid cells, and production was triggered by IL-23 production by CD103+ cDC1 (59). We could also identify that IL-22 acted on TECs to stimulate their repair, showing that TEC proliferation and survival was enhanced following IL-22 administration. However, the specific mechanisms of IL-22 on TECs are still largely undefined, although subsequent studies have suggested a potential role in regulating Foxn1 (60–62). However, only approximately one third of TECs – across cTECs and mTECs – express the IL-22 receptor, suggesting that that IL-22 only affects a subset of TECs, although the relative differences in TEC function between Il-22R+ and IL-22R− cells is not yet clear. Since then there has been considerable work to tease apart TEC heterogeneity, primarily with scRNAseq, but given its low expression these datasets have yielded considerable insight into IL-22R restriction.

Previous work by us and others has shown that the thymus is sensitive target of GVHD, and that this damage may even contribute to chronic GVHD due to a breakdown in central tolerance (63–65). Much like the gut, we were able to subsequently show that thymic ILCs were also a target during GVHD, and that this loss of IL-22, in addition to the direct depletion of TECs during GVHD, was a significant contributor to post-GVHD failure for thymic regeneration (66).

Endothelial Cells and BMP4

Although mice deficient for IL-22 knockout had impaired regeneration compared to their WT counterparts, there was still some regeneration within these mice (59). This suggested to us that there was some redundancy in thymic regeneration and there were likely alternate pathways also contributing to thymic repair. In early transcriptomic analysis of the non-hematopoietic stromal compartment of the thymus after injury (comprising TECs, but also endothelial and mesenchymal cells), we noticed that a) BMP4 was upregulated shortly after damage, at a timepoint consistent with when the thymus is starting to repair itself, and b) an enrichment for genes downstream of BMP signaling (67). This suggested an alternate pathway of thymic regeneration, centered on the production of BMP4. Expression of BMP4 was restricted to fibroblasts and endothelial cells, but only ECs increased their expression after damage suggesting that this could be the most important population for their production and its role in repair. ECs had been previously found to be resistant to injury in the thymus (68). Consistent with this, deletion of BMP4 specifically (and inducible) in ECs using the Cdh5-CreERT2 mouse led to worse thymic recovery following damage. BMP4 had previously been described in thymic function, largely during ontogeny, but notably one of its known roles was in direct induction of Foxn1 by TECs (69, 70). This was especially notable since we and others have shown that in the context of acute damage Foxn1 and its crucial downstream targets such as Dll4 and Kitl are increased. We could therefore demonstrate that upregulation of BMP4 could induce TEC expression of Foxn1 and its downstream targets (67). However, this seemed largely restricted to cTECs with little upregulation of Foxn1 in mTECs, although we did find BMP4 induced mTEC proliferation through heretofore unclear mechanisms. Subsequent analysis of BMP receptor expression suggests widespread expression across most TEC subsets (71).

Myeloid cells and Type II Immune Responses

Since these first two descriptions of the mechanism regulating endogenous thymic regeneration, several other studies have emerged highlighting alternate cells and molecules involved. Most notably are several papers describing a type 2 immune circuit involving ILC2, eosinophils, tuft cells, fibroblasts and Tregs. In this system, ILC2 production of IL-4, IL-5, IL-13 and amphiregulin (a ligand for epidermal growth factor receptor (EGFR), can contribute to thymic repair after acute injury (72–75). Consistent with this pro-regenerative function of type 2 immune responses in the thymus, we have recently shown that recirculating CD4⁺ Tregs aid thymic regeneration via their production of amphiregulin (76). Tuft cells can also produce another EGFR ligand, heparin-binding epidermal growth factor (HB-EGF), which can directly stimulate TEC growth in fetal thymic organoid culture (75). Eosinophils have also been implicated in this process, being recruited to the thymus following radiation injury by ILC2-produced IL-5, and mediate regeneration via their production of IL-4 (72, 73).

Mesenchymal cells – growth factors and chemokines

Several studies have begun to untangle the heterogeneity of the thymic mesenchymal compartment, and with this diversity of populations has emerged a key role in steady state thymic function as well as endogenous regeneration (77–81).. We have recently shown that most thymic fibroblasts enact a regeneration program after damage comprised primarily of FGFs and BMPs and that this fibroblast response was blunted in aged mice (82). Fibroblast growth factors (FGFs) are produced by mesenchymal cells and act on epithelial cells across a range of tissues, including skin, lung, bladder, gastrointestinal tract, as well as thymus (83, 84). Most prominently FGF7, also known as KGF, directly acts on FGFR2IIIb expressing TECs and support the formation of the medullary epithelium (85, 86) and as discussed above exogenous KGF treatment protects TECs in from damage, suggesting the important contribution of the molecule to restoring thymic function (86). FGF21 has also emerged as a key growth factor for thymic function in aged and following allo-HCT conditioning (87). In addition to these growth factors, thymic mesenchymal cells have also been shown to be important producers of chemokines such as CCL19, CCL21, and CCL25, which are all decreased acutely after damage in the thymus but emerging as important for endogenous thymic regeneration via their role in the importation of hematopoietic progenitors (80).

4. TRIGGERS OF REGENERATION

Abrogation of apoptosis: Activating a “dead-man’s switch”

Despite the growing identification of multiple regeneration pathways and the detailed cellular and molecular mechanisms driving their function, the fact that there were in fact multiple distinct pathways emanating from distinct initiating cell types led us to question if there could be a common trigger for these pathways. In the context of IL-22 and Il-23 production, we noticed that the presence or absence of thymocytes themselves appeared to be a catalyst for their production (59). Specifically, using knockout mice with blocks at different stages of T cell development (88), we found that even in the absence of damage, mice with blocks in T cell development up to the CD4+CD8+ double positive (DP) stage had increased levels of IL-23 and IL-22, even in absolute levels; remarkable given the severe depletion of thymic size in these strains. However, this was only true for blocks before DP stage of development, including mice deficient for Rag1^−/−, Il7^−/−, Il7ra^−/− and Tcrb^−/−. In contrast, mice that were replete with DP but had a block in SP or mTEC development (Cd40l^−/−, Ccr7^−/−, Tcra^−/−), there was no change in the production of these factors (59). This led us to hypothesize that something about the thymocytes was suppressing the production of these regenerative factors. Given that one of the most abundant features of thymocytes, and especially DP thymocytes, is that they naturally undergo apoptosis as part of positive selection (89), we, therefore, hypothesized that depletion of DP cells would lead to a lack of apoptotic thymocytes and therefore the removal of a suppression signal, initiating the production of the regenerative factors BMP4 and IL-23 (upstream of IL-22). This was supported by studies suggesting that apoptotic cells can regulate cytokine production, including the downregulation of IL-23 (90).

Indeed, in co-culture experiments we found that apoptotic thymocytes suppressed the transcription of Il23 by DCs and Bmp4 by ECs (91). Further analysis revealed that these apoptotic cells were detected by the TAM receptors, which can detect inverted phosphatidylserine, that occurs in one of the earliest stages of cell death (91). Downstream of TAM receptors we found that activation of the Rho-GTPase family (including RhoA and Rac1) were crucial for this pathway. Notably, NOD2, whose typical function is to sense bacterial peptidoglycans, has been shown to regulate IL-23 at the resolution of infection, and has also been shown to sense activated Rho-GTPases (92, 93). We found that absence of NOD2 led to enhanced thymic repair, and increased production of IL-23 and BMP4 – consistent with a role in regulating regeneration. We also found that downstream of this pathway was the microRNA29c, that inhibited translation of Bmp4 and Il-23. Interestingly, miR29 has previously been identified by the Liston group for its role in thymic involution (94) – although in our hands it was miR29c and not miR29a that mediated this pathway. This study provided the first evidence of a common regeneration regulator, implying that merely the depletion of DP thymocytes is enough to trigger the regenerative response – essentially a “dead-man’s switch” that activates on the depletion of cells, however that occurs (Figure 2).

Figure 2: Balance of cell death detection coordinates endogenous thymic regeneration.

Our work has shown that detection of modalities of cell death co-ordinate aspects of thymic regeneration. Specifically, at baseline where there is considerable apoptotic cell death during thymocyte selection (β-selection, positive selection, and negative selection), leads to the suppression of regenerative cues mediated by the detection of these apoptotic cells. However, after acute damage there is a switch toward caspase-1-mediated immunogenic cell death (ICD) that leads to the release of pro-regenerative DAMPs such as ATP and zinc. However, the activation of caspase-1 also leads to the release of pro-inflammatory IL-18 that can stimulate NK cell cytotoxicity and the depletion of thymocyte-supporting thymic epithelial cells (TECs).

Switch toward Immunogenic cell death: DAMPs, and alarmins

After damage in the thymus there is an acute depletion of cellularity that is mediated by considerable cell death. If apoptosis suppresses the production of these regenerative factors, why does this not happen in the early stages after acute damage. Firstly, there is likely a timing factor where the true regenerative program does not typically take effect until soon after the nadir of thymic cellularity. However, we wondered if there might be other explanations such as the use of alternate forms of programmed cell death (95). In fact, prior reports have suggested that damage caused by stimuli such as ionizing radiation and chemotherapy can actually induce pyroptosis, a form of immunogenic cell death (ICD) (95). While there was an increase in apoptosis after radiation injury, as measured by cleavage of caspase-3, we also found significant upregulation of caspase-1 cleavage into its active form after ionizing radiation (96). However, notably, this phenomenon was not restricted to just ionizing radiation with increased caspase-1 in other models of thymic injury as well including cyclophosphamide, and LPS as well as dexamethasone – which is considered to be a classical inducer of apoptosis (97). Early work demonstrated that apoptotic molecules such as Bax/Bak were crucial for thymocyte cell death after radiation injury (98, 99), not only was there still some cell death even in mice incapable of undergoing apoptosis, but recent evidence suggests that these same pathways can actually play a role in ICD as well as apoptosis (100). While apoptosis is generally thought to be immunologically silent, not directly inducing an immune response beyond phagocytosis, ICD through its release of intracellular contents as damage-associated molecular patterns (DAMPs), will trigger an immunological response (95). We found that in contrast to apoptotic thymocytes, pyroptotic thymocytes could directly induce FOXN1 expression in TECs, and this was mediated by the release of ATP – a prototypical DAMP – via the purinergic receptor P2Y2 (96).

Curiously, we found that the release of zinc after acute damage could also induce the production of BMP4 and IL-23 (101). This work highlighted a novel DAMP in zinc, which is the second most abundant trace element in the body and capable of interacting with approximately 300 proteins, including those involved in immune health (102, 103). We found that zinc was required for normal T cell development, helping with thymocyte expansion but also accumulating within thymocytes as they develop. After damage this stored zinc is released and is important for inducing endogenous thymic regeneration, specifically by promoting the production of BMP4 by ECs and IL-23 by DCs via the Zn-sensing G-protein coupled receptor GPR39 (104).

Consistent with this role for ICD in thymic repair, studies suggest that the alarmins IL-33 and IL-25 were crucial for activation of the regenerative pathway centered on the production of type II cytokines (72, 73, 75). However, interestingly, in this context there appears to be a key distinction in the trigger based on the type of damage. For instance, following sublethal ionizing radiation, the trigger came from fibroblast release of IL-33, whereas in the context of the corticosteroid dexamethasone, the regeneration pathway was triggered by IL-25 produced by Tuft cells (72, 73, 75). While it is not yet clear the relevance of these differential responses, it is worth noting the constitutive nature of IL-33 expression making this pro-regenerative factor poised for effector function(105). Furthermore, this fascinating distinction is one of the first instances separating different modalities of damage and will likely yield novel findings on tissue repair in the future

5. FACTORS THAT MODULATE THYMIC REGENERATIVE RESPONSES

Inflammation

Although there is growing evidence that many DAMPs can be pro-regenerative, given the pro-inflammatory nature of ICD it is perhaps not surprising that it is not as simple as that. For instance, we have recently shown that in addition to the release of the pro-regenerative DAMPs Zn and ATP, activation of caspase-1 also leads to the cleavage of pro-IL-1β and pro-IL-18 into their active forms. While IL-1β did not have any effect on thymic regeneration after injury, IL-18 suppressed TEC regeneration via stimulation and activation of NK cells and their production of perforin and Granzyme B (97). Furthermore, mice deficient for the inflammasome component NLRP3 – which is typically upstream of ICD – have increased thymic regeneration and T cell reconstitution following HCT (106). Consistent with these findings, mice deficient for NLRP3, an inflammasome component upstream of thymocyte pyroptosis, have improved thymic recovery and T cell reconstitution post-HCT, as well as decreased age-related thymus involution (106).

Graft versus host disease (GVHD)

In the context of allogeneic HCT, one of the main causes of mortality is GVHD, which can severely impact organs such as gut, liver, skin and even brain (7, 107). However, we and others have shown that the thymus is an especially sensitive target for acute GVHD, leading to direct loss of thymic epithelial cells (63, 65, 108, 109). In fact, this targeting of the TEC network, and especially the breakdown in tolerance induction that this can cause, a key trigger for the initiation of chronic GVHD, a later forming disease (63, 64, 110–112). In addition, in both gut and thymus we have shown that in addition to eliminating crucial epithelial cells in those tissues, alloreactive T cells also target regeneration-initiating cells such as ILCs, reducing their production of IL-22 and dampening regeneration (56, 66). It is likely this multipronged damage contributes to the severity of GVHD-related injury. Notably, exogenous replacement of IL-22 leads to improved function in both the gut and the thymus and, importantly, a clinical trial has shown efficacy in steroid-refractory GVHD in the gut (113). However, even though GVHD itself is severely limiting of thymic regeneration, paradoxically so to can many of the prophylactic strategies to abrogate the initiation of GVHD, including ATG, post-transplant cyclophosphamide, and tacrolimus (112, 114–119). It is perhaps for this reason that while post-transplant cyclophosphamide (PTCy) has emerged as a potent inhibitor of GVHD, it did not actually lead to an increase in overall survival in a recently published Phase 3 trial (120). Clinically, there is a strong correlation between GVHD and reduction in thymic function as measured by TRECs (121, 122). This is especially important given the increasing evidence that CD4 T cell reconstitution is predictive of overall survival following allo-HCT irrespective of donor type and across the main causes of morbidity and mortality (13, 14, 123).

Aging

Age-related thymic involution is not only a hallmark feature of aging being one of the earliest organs to show signs of age, but is remarkably conserved across species, occurring in almost all vertebrates (124, 125). Thymic involution is characterized by the progressive decline in T cell development and output (126, 127). Although age-induced reduction in lymphoid progenitors can exacerbate thymic involution (128), evidence suggests that it is alterations in the thymic microenvironment that are ultimately most responsible for driving the age-related decline in the production of T cells (129). The consequences of this decline in T cell production ultimately manifest in reduced naïve T cell output, coupled with the oligoclonal expansion of existing memory T cells, which leads to a constriction in TCR repertoire diversity and reduces the probability of a sufficient immune response (130–132). Historically it was thought that the thymus was not relevant past childhood (8, 133, 134), however, there is growing evidence that its postnatal function does indeed contribute to ongoing immune function. For instance, children who have received thymectomy (typically as part of cardiac surgery), have been found to exhibit reduced responsiveness to new vaccine (135–137), and more recently, a study of adult thymectomy recipients found an increased all-cause mortality as well as increased cancer incidence (138), strongly implicating thymic function in ongoing immune health. Ultimately, reduced thymic function, whether through age, poor regeneration, or thymectomy, leads to a constriction of the T cell receptor repertoire and reduced probability of producing a clone that will recognize the needed antigen (131).

However, in addition to these effects of age on steady state thymic function, aging also profoundly modulates endogenous regenerative responses, with vastly reduced capacity for regeneration with age (8, 114). In the context of allo-HCT, there is a well-established correlation between age and CD4⁺ T cell reconstitution (11, 12, 139, 140). This prolonged period of T cell immunodeficiency in older individuals increases vulnerability to opportunistic infections and relapse of malignancy (13, 141). However, although these phenomena of poor reconstitution in aged individuals are well-established, the causes of have been insufficiently studied. A recent paper from our groups sought to shed light on this phenomena by studying stromal responses to damage across lifespan (82). We found that not only is the regenerative response blunted with age – with reduced expression of key regenerative factors such as FGFs and BMPs from endothelial cells and fibroblasts, but that the emergence of non-productive age-associated TECs (aaTECs) with age blocks recovery of conventional TECs by stealing regenerative signals after damage (82).

6. CONSEQUENCES OF PROLONGED DEPLETION OF THYMIC FUNCTION

There is a clear need for thymic function in the context of allo-HCT to generate new T cells, with the only other source of reconstitution coming from T cells in the graft. As such a lot of the evidence for the consequences come from allo-HCT. For instance, as discussed earlier, there is a growing body of evidence that suggests CD4 T cell reconstitution can be predictive of clinical outcomes across a range of transplant complications, including relapse, GVHD and infection (13, 14, 16). While not addressing effects after damage per se, there is also emerging evidence that poor thymic function in general can also be consequential for health outcomes, largely as poor T cell reconstitution ultimately leads to prolonged constriction of the TCR repertoire, an outcome similar to that of age-related thymic involution. Relatedly, there is evidence that the effects of acute damage can lead to long-term functional deficits, highlighted by immune defects in the survivors of the nuclear bombs in Hiroshima and Nagasaki (142, 143).

The functional consequences of reduced thymic function were especially highlighted during COVID-19: constriction of the TCR repertoire in patients over 50 was linked with an increased risk of severe illness and reduced vaccine responsiveness (144–147). Additionally, there is also mounting evidence that thymic function can contribute to cancer outcomes, with the aforementioned study linking thymectomy to increased all-cause mortality and cancer incidence (138). This finding is consistent with the potential reduction in cancer immunosurveillance upon TCR repertoire constriction, as well as the strong correlation between the mutational burden of the tumor and responsiveness to immune checkpoint inhibition (131, 148–151). This is because in all of these instances, responses are predicated on the presence of a T cell clone that recognizes the tumor neoantigen, pathogen or vaccine antigen; the probability of which declines in parallel with the TCR repertoire constriction (131). Consequently, TCR repertoire breadth has been proposed as a biomarker for anti-tumor responses (152–155).

7. RISKS OF BOOSTING THYMIC FUCNTION

While there is considerable promise for improving T cell recovery in contexts such as HCT, where prolonged suppression of adaptive immunity can contribute significantly to transplant outcomes, there are some potential risks that may need to be considered when thinking of applying these therapies beyond transplant. For instance, in the context of age, where the desired outcome of boosting thymic function would be to broaden the TCR repertoire, it is worth considering how processes like central tolerance will be affected. In fact, in one study it was proposed that while sex steroid inhibition-mediated thymic regeneration did enhance the size of the organ, many of the functional changes with age persisted, suggesting that there could be significant implications for tolerance and autoimmunity (156).

8. CONCLUSION

The thymus is crucial for generating a diverse and self-tolerant T cell receptor repertoire, a process that relies on endogenous tissue regeneration in response to a range of acute and chronic stressors. This is especially important given the growing body of evidence that postnatal thymic function is also crucial for maintaining this broad repertoire of T cells to ensure recognition of unknown pathogenic antigens, vaccines, or cancer neoantigens. While this dynamic function of the thymus has been noted for almost as long as its existence, we and others have only begun to elucidate the specific molecular and cellular mechanisms that underpin endogenous regeneration as well as the triggers that activate these pathways. Although it will be important to establish the translatability of these pathways to human tissue - which has been traditionally limited to available, primarily pediatric, thymic tissue - there is considerable potential in these regenerative pathways for improving thymic function and immune health.